he West Indies, follow
the courses where the crust of the earth is thinnest and where great
bodies of water lie on the shallowest parts of the ocean bed.

The terrific heat of the earth's internal fires is sufficient to cause
crevices leading from these bodies of water to the central fires of
the volcano, and the character of the volcanic eruption is determined
largely by the size of the crevices so created and the amount of water
which finds its way through them. The temperature of these internal
fires can only be guessed at, but some idea may be formed of their
intense heat from the streams of lava emitted from the volcano. These
will sometimes run ten or twelve miles in the open air before cooling
sufficiently to solidify. From this it will be seen that the fires are
much hotter than are required merely to reduce the rock to a liquid
form. From this fact, too, may be seen the instantaneous action by
which the water seeping or flowing into the volcano's heart is
converted into steam and a tremendous explosive power generated.

The calamity which befell Martinique and St. Vincent will
unquestionably lead to a fresh discussion of the causes of volcanic
disturbance. Not all of the phenomena involved therein are yet fully
understood, and concerning some of them there are perceptible
differences of opinion among experts. On at least one point, however,
there is general agreement. At a depth of about thirty miles the
internal heat of the earth is probably great enough to melt every
known substance. Confinement may keep in a rigid condition the
material which lies beneath the solid crust, but if an avenue of
escape is once opened the stuff would soften and ooze upward. There is
a growing tendency, moreover, to recognize the importance of
gravitation in producing eruptions. The weight of several miles of
rock is almost inconceivable, and it certainly ought to compel
"potentially plastic" matter to rise through any crevice that might
be newly formed. Russell, Gilbert and some other authorities regard
this as the chief mechanical agent in an eruption, at least when there
is a considerable outpouring of lava.

As to the extent to which water operates there is some lack of harmony
among volcanists. Shaler, Milne and others hold that substance
largely, if not entirely, responsible for the trouble. They point to
the fact that many volcanoes are situated near the coast of continents
or on islands, where leakage from the ocean may